how artificial intelligence will affect law firms?

O’Brien 1

How Artificial Intelligence Will Affect the Practice of Law?

Cody O’Brien

HN-300-A: Honors Capstone

Dr. Andrew Lazowski; Dr. Daniel Rober

Spring 2019

1

O'Brien: How Artificial Intelligence Will Affect Law Firms?

Published by DigitalCommons@SHU, 2019

O’Brien 2

Abstract

This paper aims to give a fundamental understanding of how artificial intelligence will

affect the legal industry. The overall purpose of this study is to explore a wide range of topics

that ultimately show how AI can be most effective in our workforce. Currently, AI has many

setbacks in terms of its logical-decision making, meaning, when AI is used in our legal system,

there is always legal precedence that should never be overlooked. The basic design of the study

is discussing innovative AI programs, legal profession v. human intelligence, ethical aspects, law

school curriculum, and the global job market at law firms. I focused more on the domestic legal

system in my argument, but I also touched on international statistics to help shape a broader

macro outlook. As a result, we will see a multifaceted legal industry that is interdependent in

many ways. Furthermore, the U.S. legal system is looking to continually improve the way AI is

implemented at law firms and in court while learning how it reacts in a different geopolitical

environment. In summary, my research analysis will explain how AI is useful to many law firms,

but more importantly, recognize the fact that it lacks “human individuality” in the way

information is being presented.

2

Academic Festival, Event 50 [2019]

https://digitalcommons.sacredheart.edu/acadfest/2019/all/50

O’Brien 3

Introduction:

Artificial Intelligence is a notoriously difficult term to define. By definition, “AI is the

theory and development of computer systems able to perform tasks that normally require human

intelligence, such as visual perception, speech recognition, decision-making, and translation

between languages” (English Oxford Dictionary). AI is becoming more and more intertwined

with our workforce, especially in the practice of law. Artificial intelligence has become very

prominent within law practices and touch on instances that give us a futuristic outlook in the

legal industry. Today, AI is being utilized to eliminate many tasks through automation,

programming, and mechanization at law firms. However, my research will discover whether AI

can efficiently and legally replace many of these middle-to-back office tasks and help lawyers

lower their total volume of responsibilities. In many ways, AI is supportive in the decision-

making process for lawyers. Much legal criteria in cases require lawyers to develop favorable

and cost-effective strategies for clients, prepare legal documents, and stay informed about current

legislation, especially concerning legal specialties. AI can be tremendously beneficial and cost-

effective for growing law firms.

Much of the time when lawyers lose, it is because the other one is smarter. Judgment

plays a vital role in court cases on both sides. Now, one question presented to developing law

firms is figuring out how AI can be implemented to their business model. Nonetheless, firms

need to consider that AI has certain precedence over logical decision making. There are common

legal procedures that need to be done before utilizing automated programs. For instance, when

attorneys build relationships with clients, they must establish trust more than anything. No

artificial intelligence worldwide can uphold healthy and successful relationships than a human

being develops through social interactions. Instead of having clients provide uncertain truth,

3



O’Brien 4

cases will now become much smoother since both sides have a solid story to confidently argue.

Law firms are continually looking for ways AI can be most beneficial over time, but certainly

these technological advancements will be required to evolve over time.

Section 1 – Innovation

Law firms are consistently looking for ways to innovate while being efficient throughout

the process. The business model for many law firms today does not require legal education but

mere data procession. Certainly, this does not apply to all employees at a law firm, but firms are

transitioning to a pure data-oriented model. Two law firms have made significant progress with

automated technology that works with these firms. In 2016, Linklaters and Pinsent Masons

announced their decision to invest in autonomous office automations. Linklaters had developed

“Verifi”, which is a computer program that can sift through 14 UK and European regulatory

registers to check client names for banks. Verifi has the capability to process thousands of names

overnight. Law firms would save a considerable amount of time using Verifi as opposed to

tracking down client’s names or accounts across the globe. The latest Binder Dijker Otte (BDO)

law firm leadership survey states that, “Artificial intelligence will have the greatest impact, with

many believing it would replace the work of lawyers or strip out a significant layer of work and

revenue from law firms” (BDO 2017). This will, in turn, restructure business models and

finances within law firms.

Although this automated technology can eliminate many office tasks, there is a

substantial amount of work that needs to be accomplished through a lawyer. Sabrina Praduroux

is an author and researcher who lays out several categories in which AI technology can work

together with law firms to ultimately help drive success.

Section 2 – Current AI Systems

4



O’Brien 5

The first is called Lawyer-to-Lawyer Networks which provide synergies for outsourcing.

In essence, this network enables law firms around the world to create social and referral

networks. Using lawyer-to-lawyer networks can be extremely useful when one lawyer needs to

get information from another like-minded lawyer. For example, suppose that a business lawyer

was drafting a contract for a venue (i.e. hotel) and needed to generate an in-depth analysis of the

total square footage and financials. Now, by using this type of network, that lawyer can get in

touch with another lawyer at a similar law firm who knows the total square footage area as well

as the creditworthiness for this facility. This is a very exciting development that many law firms

could benefit from using. Lawyer-to-Lawyer Networks will make the drafting process much less

tedious, while also creating close relationships with other firms.

Secondly, Document Automation and Assembly, encompasses the design of systems and

workflows that assist in the creation of electronic documents. This is a crucial task that law firms

will always need to get done. These include logic-based systems that use segments of pre-

existing text or data to assemble a new document. This technology will be used for numerous

purposes including, but not limited to, analyzing contracts, comparing Non-Disclosure

Agreements, generating memo’s, and presentations. In addition, it could also include the so-

called ‘smart contracting’ which refrains from controlling actions of people and non-corporeal

entities in an environment that is basically borderless. The idea of smart contracting serves to shy

away from actions of people in cases, since that argument can be very wide open in court.

Thirdly, Practice Management is a practice and case management software that provides

attorneys with convenient methods for effectively managing client and case information,

including contracts, calendar and meeting information, documents, and other specifics. In my

business internship last summer, I was able to compare two documents in a word document at a

5



O’Brien 6

basic level. Practice management is designed to organize client data such that lawyers can access

their information much quicker. Case management software would also schedule client meetings

and systematically list all contracts and meeting information involved. Although this software

might not be applicable to every law firm’s business model, it can certainly be used to get

multiple tasks done at a faster rate.

Fourthly, Predictive Analytics and Litigation Data Mining is an analysis of data through

statistical or mathematical techniques that results in meaningful relationships being identified in

data. One example, is if a lawyer was looking for significant phrase or phrases in a document;

predictive analytics works to make it clear that both statements correlate with each other. The

results can then be used for better prediction of future events and better decision-making in each

case. A lawyer can also filter out phrases within a contract to decide how meaningful their

argument is moving forward. Predictive modeling of litigation management provides information

needed at the beginning of a judicial process to improve it. This will ultimately provide lawyers

with resources needed to prepare for court way in advance.

Although these technological advancements are supportive for law firms, there are

several problems of practical implementation. For instance, “hard cases” are cases that entail

reasoning influenced by extra-legal elements, and could very well be “fact determiners” when

processing legal documents. In other words, hard cases do not represent a solid base for general

law. For example, let’s assume a lawyer had to defend a serial killer on trial. In this case, it

would be more beneficial if that lawyer proposed facts and laws by himself rather than having an

automated system weigh this case. The defense can now make a more effective argument that is

accurately tailored to their side of the story. In a globalized world, it is also important to

6



O’Brien 7

recognize that potential uniform automatization of law depends on the geographical location

because laws vary substantially from one jurisdiction to the next.

Section 3 – Legal Profession

The role of technology at law firms is multifaceted. There are numerous innovative and

efficient methods to outperform manpower typically needed. In The Legal Profession: From

Humans to Robots, Jordan Bigda explains various reasons why robots and artificially intelligent

programs can very well be the future of law. Bigda puts forth three impacts of how AI affects

law firms which show whether or not this system can be useful.

First, she explains how these technological processes provide lean and cost-efficient legal

services. Secondly, we tend to see artificial intelligence only performing associate level work.

Thirdly, a proceeding challenge for this major transformation is whether firms should no longer

pay for associates or paralegals. Ultimately, there are pros and cons at law firms, but it is more

important to know that law firms today are continually learning where AI can be most useful.

If we have no more graduating legal associates, then who becomes lawyers of the future?

When more and more college students see that legal associates are being replaced by AI, it can

ultimately dictate whether they choose that field of study. Of course, a career path for someone

to become a lawyer requires some level of experience as a paralegal. Now when AI transforms

our job market, it can be detrimental to the overall economy.

With AI developing at law firms, it is possible that there will be new and different

employment opportunities for lawyers going forward. Robots and programmers are not lawyers;

therefore, the use of robots is limited by law. In The Legal Profession, Bigda considers

automated systems as the “unauthorized practice of law” (Bigda, ‘The Legal Profession’). In

some respects, there are ways AI can work more efficiently, but certainly, the question raised to

7



O’Brien 8

this process is how conventional it would be. During the 1980’s, we saw a large increase in

lawyers being hired, but, by the 1990’s law firms had to find ways to cut costs. Astonishingly,

libraries of books had been replaced by computers.

Section 4 – Legal Industry

Tasks of paralegals and legal assistants include, researching cases, preparing discovery

law in common jurisdictions, interviewing clients/witnesses, generating case summaries and

general case management. However, due to court decisions, paralegals cannot advise clients,

mediate, or negotiate on behalf of clients. With the introduction of computers – law firms,

offices, and staff become smaller. There are already programs that create documents for people,

meanwhile, clients will no longer have to pay a lawyer to fill out paperwork which saves a

tremendous amount of everyone’s time. But, a huge question of whether or not automated

systems carry out this so-called “unauthorized practice of law” is shaping up.

“Kira” is a form of AI for law – used for contract review, analysis, or knowledgeable

management. Kira can mitigate the risk of errors, enable faster deal-making, increase speed, and

ultimately, improve value while keeping both the client and lawyer satisfied. Firms can use Kira

to get a solid framework of facts in a case much quicker. This would enhance the deal-making

process on both sides of a case.

“Ross” is characterized as an artificially intelligent lawyer that assists human lawyers in

research. Now, lawyers can focus on advising clients. Ross is based in IBM’s “Watson”

technology, meaning its capable of answering questions posed in natural language. One benefit

of Ross is it has the ability to update its work and learn new material presented. When a lawyer

needs to ensure that new case material is being updated, Watson would essentially correct the

information instantaneously.

8



O’Brien 9

Although change is about to happen, law firms should prepare and support the use of AI

for their client’s sake. AI does not have the ability to apply new law to cases. However, with

Ross and Kira it becomes possible to efficiently replace a portion of the work. First, you program

these technology systems with legal information, including statues and all case law. Keep in

mind that lawyers need to program “Kira and Ross” as they work. This allows these robotic

systems to continually become more knowledgeable in their field. If AI programs are able to

negotiate, mediate, and advise clients, then this would be the performance of a legal task and

lawyers would not be necessary. New law must be created to overcome these ethical issues on

the brink to controversy. Whenever there is a significant change in a market or profession,

regulations and rules should accompany these changes.

Section 5 – AI’s Impact at Law Firms

In How AI will Affect the Practice of Law, we see that new software tools present both

new challenges and new opportunities. AI has a profound impact on law firms and many of their

operating procedures. Near term, we can expect greater legal transparency, more efficient dispute

resolution, improved access to justice, and new challenges to the traditional organization of

private law firms. These new challenges include delivering legal services on a billable hour basis

through a leveraged partner-associate model. In other words, instead of paying a lawyer and his

or her staff on an hourly basis, AI now has a place to save time and money.

What do lawyers do? We typically think of lawyers as highly trained and highly skilled

professionals who identify legal issues, gather relevant facts, and determine the likely outcome

of a case. In addition, lawyers exercise judgment, using their experience and intuition to

determine the best way to proceed. Given the importance of expert human judgment, one might

think the cost of a lawyer’s service would only increase over time. Recent development in

9



O’Brien 10

artificial intelligence, in particular, natural language processing and machine learning have

challenged traditional conceptions of human expertise.

Machines now perform increasingly complex tasks much better than humans. Tasks that

used to require human effort, such as complex calculations, counting money, creating

spreadsheets, conducting risk assessments are now transitioning into automated roles that not

only reduce cost but also offer greater accuracy and precision. “John McGinnis and Russell

Pearce have argued that machine intelligence will cause a ‘great disruption’ in the market for

legal services” (McGinnis and Pearce, ‘Great Disruption’). Both of them specifically pointed out

five areas of legal services to be affected are discovery, legal search, document generation, brief

generation, and prediction of outcomes.

Machine learning has potential to provide more objective predictions of how courts will

decide discrete legal issues. Once given facts relevant to a question, a machine can then situate

these facts within the domain of applicable legal precedents. In other words, machine learning

can gather factual information with relevant legal precedence. Additionally, algorithms can

generate predictions that other firms can replicate. More and more firms will now be able to let

computers do the majority of work since algorithms essentially make predictive analyses. Such

technological advances make it possible for parties to litigate more effectively before, during,

and after trial, by predicting legal outcomes with greater accuracy and consistency at a lower

cost.

Section 6 – Practical Effects at Law Firms

After the 2008 recession, lawyers across diverse practice settings have explored ways to

provide more practical legal representation. This economic downturn gave clients of law firms

leverage to demand cost-effective representation. Law firms began experimenting with

10



O’Brien 11

alternative approaches such as flat rates and contingency fees which were less expensive for

clients than hourly billing.

Artificial intelligence played a substantial role at the time. Post-recession, development

and use of automated systems has never been more prevalent. Situations that give rise to legal

representation are rarely answered definitively by existing precedent or statues. In other words,

not every situation is going to have the same facts in a case which follow that specific legal

precedence. Every day laws are constantly evolving and lawyers must draw analogies and

identify distinctions from relevant cases to best promote their clients’ interests.

Legal scholars have suggested that because lawyers self-regulate their own market, they

will be able to offer strong resistance to development, use, and integration of machines into legal

practice. We notice firsthand today how each lawyer will ultimately tailor their work processes

how they see fit. Therefore, a technological advancement will not setback lawyers work, but

rather provide efficient methods to get the job done. Natural language processing enables

machines to adapt when evaluating text. This could be very helpful for lawyers to read and

interpret long documents. As a result, “lawyers now spend less than 5% of their time on basic

document review” (Remus and Levy, ‘Can Robots be Lawyers’). A central challenge for using

AI in law is the unstructured nature of legal data. For example, judicial opinions follow a general

form – recitation of facts, discussion of relevant case law, and application of law to the facts,

however, judges are highly individualistic in the way lawyers present this information. When

lawyers can verbally articulate their argument, it becomes more appealing to the judge and

makes them sound more authentic.

Section 7 – AI and Knowledge Management (KM)

11



O’Brien 12

Another essential article, Artificial Intelligence and Knowledge Management:

Questioning the Tacit Dimension, Louis Sanzogni researched developing AI to matching its

capabilities to the human mind. Although Sanzogni concluded that right now it is a work in

progress, he looks at the human mind from a unique angle and its relation to AI. AI is still

focused on automating what is largely verbally articulated.

Artificial agents operate in what is referred to as an ‘objective reality’ where, at least in

principle, its aspect should be verifiable. In other words, artificial agents are on track to basically

revise and verify the automated work produced. Sanzogni takes on an interesting perspective of

the mind and artificially intelligent programs moving forward. In reality, there is a fine line with

human thinking and AI. Recognizing, for example, methodological disagreements between rule-

based and neural computational models is open to possibly having cognitively-aware machines

going forward. In simpler terms, rule-based systems are used to interpret information in a useful

way whereas a neural model is based on recognition. As a result of this distinction, there is a

probability that we will see cognitive functioning in machines in the legal industry.

Knowledge has different meanings and roles depending on the perspective taken. Tacit

knowledge, which in essence is implied knowledge, can be demonstrated in actions of practice

and doing. Knowledge is also seen as relational since its mediated through artefacts that might

have diverse logics of actions and history. Some, but not all, relational tacit knowledge can

extract corporate social responsibility values from company documents and match these values

to the financial outcomes of a company. All AI can do in practice is store articulated rules and

apply these to increasingly complicated situations. The current goal of AI appears unreachable as

computers lack not only self-awareness, but properties acquired by humans in a social

environment.

12



O’Brien 13

Section 8 – Tacit Knowledge Example

In one example, consider programming tacit knowledge, specifically how something is

being said. If we look at sense making, meaning, wisdom, or emotions, there is a unique way it

can be communicated. Assume that two people off the street tried to explain my BIO-101

research paper I completed to me. Consider one person who is a doctor and the other who was a

high school drop-out with not much biology knowledge. We would see two different ways of

knowledge explained. The doctor would most likely be able to articulate the ‘meaning’ out of the

paper as opposed to the inexperienced drop-out. In this situation, a computer would be capable of

making sense of the paper regardless of outside factors that could put it at a disadvantage. The

question presented to many legal researchers is what would happen as the next generation of

street-smart changes. In other words, legal researches continue to analyze how tacit knowledge

can present information in an environment that is constantly evolving.

Computers are unable to reflect on their own performance as compared to a human

person. For instance, humans will not forgive computer mistakes, especially when money is

involved. Assume Company ABC had a lawyer draft a contract using AI, which, in turn, agreed

to have Company XYZ pay $1,000,000 to them over the next 5 years. In the contract that

Company XYZ received, there was a miscalculated amount in the yearly billing payment section

because of an AI error. Now, come time to pay this bill at the end of the year, Company ABC

lets Company XYZ know that you actually owe double that amount. Company XYZ now argues

that the contract stated one amount and you are telling me that I owe double. The moral of this

story shows that although AI can have conflicting presence in the legal profession, it is

ultimately up to the companies to check the documents either in person or through their own AI

to verify the validity.

13



O’Brien 14

Computers can never step outside code, reflect on code, and contribute their own

observations. To date, there is no AI machine able to ‘learn’ collective tacit knowledge. Like I

said, there is a distinct difference between AI and implied knowledge. The way AI and

knowledge management evolve is related to how society is organized and technological

development trends. In Artificial Intelligence and Knowledge Management: Questioning the

Tacit Dimension, Sanzogni states “AI and KM developments are a function of technology

developers’ goals and assumptions about how the world works” (Sanzogni, 50). AI and KM is

only as good as the developers’ desire to create new systematic programs. One challenge of AI is

to understand and code the analogue expression of human speech. The digital facts are easy to

express and being coded right now, however, if AI can express legal information in the same

way as a human speaking, that would be a major breakthrough.

Section 9 – Ethical Components

In Ethical Implications of Electronic Communication and Storage of Client Information,

Drew Simshaw describes why law firms truly believe new technology can negatively impact

their firm. As the ABA Cybersecurity Handbook explains, “creating, using, communicating and

storing information in electronic form greatly increases potential for unauthorized access, use,

disclosure, and alteration, as well as the risk of loss or destruction” (Rhodes and Polley, The

ABA Cybersecurity Handbook). With new technologies, comes new risk that can threaten

client’s information, and consequently, a lawyers good standing and reputation. Given the

growing threat of cyber-attacks at law firms, a lawyer’s reputation has become a necessity for

every client’s decision to hire that individual. Lawyers and law offices are facing unprecedented

challenges from widespread use of electronic records and mobile devices. Lawyers have become

targets because they collect and store large amounts of critical, highly valuable corporate

14



O’Brien 15

records, including intellectual property, strategic business data, and litigation-related theories.

Not only are law firms held accountable for their services, lawyers are acting as a middleman

between client’s and law firm’s to essentially uphold ethical standards and practices.

Section 10 – A Lawyers Ethical Responsibility

Lawyers are required to remain competent when using this technology. It also requires

continued vigilance and learning as technology advances, in order to comply with a lawyer’s

duties under ethics rules. Confidentiality is a serious matter on both sides of a case per se. First

of all, there should be a confidentiality agreement set out when lawyers electronically send

stored information to a client. Ultimately, this will ensure that both the lawyer and client

recognize if there was ever a cyberattack, lawsuits against one another would be forbidden. This

situation also applies to each side in a case. As a result, confidentiality agreements would entail

numerous risks that may arise if the defense sent information in electronic form to the

prosecution.

Attorneys must implement administrative, technical, and physical safeguards to meet

their obligation and make reasonable efforts to proceed with client information. Firms should

carefully select and supervise third-party vendors who deal with client information. Many data

security practices of third-party vendors are to ensure they are consistent with a lawyer’s ethical

obligation. Dealing with third-party vendors must be taken into account with a lot of precaution.

For example, cloud computing has become a concerning topic among law firms today. The ABA

defines cloud computing as, “Any system whereby a lawyer stores digital information on servers

or systems that are not under close control of the lawyer or law firms” (Handbook, 77). In other

words, lawyers who use traditional methods to store information on computers is now becoming

15



O’Brien 16

questionable for clients. Law firms are recommended to have private and protected servers that

will lower the risk of cyber-attacks.

Ethical rules also require lawyers to properly supervise those who work for them, both

inside and outside law firms. In other words, any legal assistant or mid to lower-level worker

must comply with similar ethical guidelines presented to lawyers.

Social media has become a very controversial topic at law firms in many ways. The ABA

Cybersecurity Handbook states, “Lawyers who share information on Facebook, Twitter, or

LinkedIn could portray blurred distinctions between a lawyers personal or professional activity

which opens the possibility that client confidential information may be shared without careful

deliberation” (Handbook, 78-79). In essence, there are multitude of cases where a legal

professional will publicly share information on social media causing unintended consequences.

A lawyer could have unknowingly posted something minor about a recent case to friends on

Facebook which, in turn, could hold him or her accountable for negligent care of confidential

information. This example becomes messy on both the client and lawyers’ side. On one hand, the

client could argue that their information was clearly exposed to the public without full

disclosure. However, the lawyer could push-back and say, “How do we measure the severity or

impact that this confidential information really had when posted on Facebook?” All-in-all, I

would say it is best for attorneys and lawyers to stay far away from that line of sharing any

relevant information on social media.

Section 10 – Revamping Law School Curriculum?

Should law school curriculum prepare students for an automated industry such as AI? AI

is only as good as those people entering the data. Aside from replacing numerous administrative

tasks through AI, there is almost always going to be a demand for lawyers to oversee these

16



O’Brien 17

processes. If young lawyers are perceived as replaceable, what will happen in the future when

there will be no more experienced lawyers. McKinsey and Company found that “23% of the

average attorney’s job can be replaced by robots or AI. It is estimated we are at 13% now”

(Johnson, ‘Find Out If a Robot Will Take Your Job’). Do lawyers have anything to fear? Deloitte

estimates “100,000 legal jobs will be eliminated by automation in the United Kingdom by 2025”

(Deloitte Insight, 2016). In addition, “J.P. Morgan used an AI computed program to replace

360,000 billable hours of attorney work” (Bloomberg, 2017). As time passes, AI in the practice

of law is seen as an opportunity rather than a threat. As with any profession and new technology,

AI is changing the way lawyers think and the way they do business.

Section 11 – Invaluable Skills of Legal Professionals

Important skills based on human judgment such as inference, common sense,

interpersonal skills, and experience, will remain valuable for the lifetime of any lawyer

practicing today. The two most important attributes for legal applications are machine learning

where AI can learn from experience to improve its capabilities, and natural language processing

which is capable of understanding the meaning of spoken or written speech.

Law schools should recognize that AI is changing the subject matter of what lawyers

work on. One example is Technology-Assisted Review (TAR), which can organize, analyze, and

search large data sets for lawyers. A second example is legal analytics which uses data,

algorithms, and AI to make predictions and detect trends (i.e. determining outcomes in

intellectual property litigation). Thirdly, legal bots are interactive online programs that provide

customized answers to specific legal situations. Given all these advancements, law schools must

still teach fundamental legal theories in order to stabilize this industry. In basic terms, aside of

where the legal profession is headed, lawyers must still acquire knowledge and laws in case all

17



O’Brien 18

else fails. I believe law schools should look at this curriculum as a win-win. If you advocate

attending a law school, then a lawyer comes out knowing a great deal of knowledge to help him

or her in their career. With AI, lawyers would benefit at work because they know what these

processes actually mean in context, however, this may require additional classes using this

technology.

Section 12 – Global Job Market at Law Firms

Regardless of where you stood in law school, increasing legal technology is changing law

practices. In Legal Education in the Blockchain Revolution, “law firms and legal tech startups are

changing how clients get involved in their cases” (Georgetown Law Center, 2016). Law firms

are dependent on individual lawyer to research your case whereas legal tech is dependent on a

computer algorithm. For example, say a client is facing bankruptcy in one state and may own

property in another. Legal tech knows that some property in specific states is protected from

bankruptcy court. Since that person is protected from out-of-state property rights, they could be

exempted from the debt payment in Chapter 7, Title 11 of the United States Law. We can see

how automated systems may not necessarily work hand-in-hand with legal tech at law firms.

A lawyer’s key skill sets are in compatible with the 21st century demand. We will begin

to see changes in legal tech that will have considerable implications for 21st century lawyers.

Legal tech is in its infancy. Prior to 2010, legal tech helped maintain law offices. Post 2010, legal

tech assisted legal professionals in due-diligence and e-discovery. Presently, “legal tech” is not

replacing lawyers. Legal tech is finding cost-efficient ways a client will get support. Legal tech

can also guide a client to prestigious lawyers or evaluate cases such that a client can decide

whether or not to move forward. This is very helpful for clients to see whether or not they should

invest in a case. Not only is this a major savings for client’s, but this process is essentially saving

18



O’Brien 19

the hassle for many firms. Law firms are being transformed into online firms, which makes sense

when you think about what legal tech can provide as a service. Legal tech will save time for

more and more clients because it would make logical-decisions throughout judicial processes.

Instead of having clients come into a law office to review details of a case, a system will now

decide whether or not to continue the judicial process. A lawyer’s time will soon become a

competitive business, ultimately trying to save client’s as much money as possible.

As we educate 21st century lawyer’s, bear in mind on how to help clients use current 21st

century technology. Today, lawyers cannot compete with legal tech, and will basically be forced

to work with this technology in order to progress.

Section 13 – Wisconsin v. Eric Loomis

➢ Case Background

In Algorithmic Injustice: How the Wisconsin Supreme Court failed to protect Due

Process rights in State v. Loomis, Katherine Freeman explains several flaws using a risk-

assessment algorithm. “Risk-assessment algorithms are equations designed to take large amounts

of information about an offender’s past criminal experience, and use it to compute a score that

ranks an offender at various risk levels” (Freeman, 75). In State v. Loomis, the Wisconsin

Supreme Court ruled that the risk assessment algorithm, entitled “Correctional Offender

Management Profiling for Alternative Sanctions (“COMPAS”) could be used for sentencing,

however, many argue that judges rely too heavily on those results during sentencing” (Freeman,

75). COMPAS was a decision support tool developed by Northpointe (now Equivant) used by

U.S. courts to access the likelihood of a defendant becoming a convicted criminal who reoffends.

➢ Case Argument

19



O’Brien 20

This recent development argues that the court misapplied precedence and offered no

actual protections of due process clause. Higher law officials proposed more effective solutions

to address the issues of COMPAS in the sentencing process. At the National Association of

Criminal Defense Lawyers – 57th Annual Meeting in 2014, criticisms of risk-assessment

algorithms arose. Attorney General Eric Holder voices his concerns. He points out how

“COMPAS undermines our efforts to ensure individualized and equal justice, and the lack of

accuracy provided” (Holder, 83).

➢ Loomis Argument

On July 13th 2016, Wisconsin Supreme Court made the decision to address

constitutionality of using risk-assessment algorithms during sentencing. Loomis presented three

due process arguments against the use of COMPAS during sentencing (Loomis, 91). First, it

violates a defendant’s right to be sentenced based upon accurate information. Secondly, it

violates a defendant’s right to an individualized sentence. Thirdly, it improperly uses gendered

assessment in sentencing. A major pushback of COMPAS is that companies hide details of their

algorithm by declaring that it is a “core piece of their business.”

➢ Solution

There are two potential plans to safeguard due process. The first is working with

companies to switch COMPAS from a proprietary algorithm to an open source algorithm that

defendants can investigate for themselves. The second is to arrange an auditing process where an

overseer from outside the company would perform consistent validation checks on the system to

ensure its accuracy and appropriate use. The result of implementing COMPAS at law firms can

20



O’Brien 21

certainly backfire, but I think with the right supervision and protection of this algorithm, it can

help make legal decisions both inside and outside of court.

Concluding Argument:

One thing is certain, there will be winners and losers among lawyers who do and do not

uptake AI. This research provides a great look of how AI can benefit law firms, but without

doubt, there are a multitude of outside risks. AI should continue to stay with law firms, but only

under the contingency that it aligns well with our United States legal system.

In Wisconsin v. Loomis, we see a perfect example of how AI is disruptive in the legal

industry. The case shows that an algorithm entitled COMPAS can be problematic because courts

tend to misapply precedence. COMPAS is useful, but needs to be accurate and appropriately

used. This algorithm is only good to a certain degree. Each case has specific legal precedent,

which is why COMPAS may not equally weigh all the evidence. Although this case presents

bias, AI can certainly save an immense amount of time and money for law firms.

In addition, plenty of innovative systems such as Lawyer-to-Lawyer Networks,

Document Automation, and Predictive Analytics have helped take a great deal of work off

lawyers. To a large extent, AI has a rippling effect on our job market and society. Replacing a lot

of these middle to back office tasks means less jobs for students coming out of law school. We

will also notice a lot of people questioning ethical responsibility of AI among law firms in

society. As discussed in my research, sending legal documents in electronic communication form

is open to cyber-attacks. This is not to say law firms are always subject to these threats, but when

handling confidential and personal information it becomes an ethical responsibility issue. If any

client private information was exposed to other parties involved, then people would question if

sending forms electronically is right to do. Ultimately, AI’s impact at law firms is a work in

21



O’Brien 22

progress. AI will continue to be present at law firms, but all these technological advancements

will be required to evolve over time with respect to each legal system jurisdiction across the

globe.

Works Cited

Bigda, Jordan. "THE LEGAL PROFESSION: FROM HUMANS TO ROBOTS." The

22



O’Brien 23

Journal of High Technology Law, Jan. 2018, p. 396+. Academic OneFile,

http://link.galegroup.com.sacredheart.idm.oclc.org. Accessed 24 Feb. 2019.

Kerikmäe, Tanel, Thomas Hoffmann, and Archil Chochia. "Legal Technology for Law

Firms: Determining Roadmaps for Innovation". Croatian International Relations Review

24.81: 91-112. https://doi.org/10.2478/cirr-2018-0005 Web.

Sanzogni, Louis, et al. “Artificial Intelligence and Knowledge Management: Questioning

the Tactic Dimension.” Prometheus, vol. 35, no. 1, Mar. 2017, pp. 37-56. EBSCOhost,

doi:10.1080/08109028.2017.1364547

Alarie, Benjamin, et al. “How Artificial Intelligence Will Affect the Practice of

Law.” University of Toronto Law Journal, vol. 68, Jan. 2018, pp. 106–124. EBSCOhost,

doi:10.3138/utlj.2017-0052.

Marchant, Gary E. "ARTIFICIAL INTELLIGENCE AND THE FUTURE OF LEGAL

PRACTICE." The SciTech Lawyer, Fall 2017, p. 20+. Academic OneFile,


Simshaw, Drew. “Ethical Implications of Electronic Communication and Storage of Client

Information.” Computer & Internet Lawyer, vol. 33, no. 8, Aug. 2016, pp. 14–

18. EBSCOhost, sacredheart.idm.oclc.org/login?url=https://search.ebscohost.com.

Fenwick, Mark, et al. "Legal Education in the Blockchain Revolution." Vanderbilt Journal

of Entertainment and Technology Law, vol. 20, no. 2, 2018, p. 351+. Academic OneFile,


Freeman, Katherine. “Algorithmic Injustice: How the Wisconsin Supreme Court failed to

23



https://doi.org/10.2478/cirr-2018-0005

https://doi.org/10.2478/cirr-2018-0005

O’Brien 24

protect Due Process rights in State v. Eric Loomis” North Carolina Journal of Law &

Technology, Volume 18. Issue on: December 2016. http://ncjolt.org/wp-

content/uploads/2016/12/Freeman_Final.pdf

Betterteam, “Attorney Job Description.” March 21st, 2018.

https://www.betterteam.com/attorney-job-description.

24



http://ncjolt.org/wp-content/uploads/2016/12/Freeman_Final.pdf

http://ncjolt.org/wp-content/uploads/2016/12/Freeman_Final.pdf

https://www.betterteam.com/attorney-job-description

https://www.betterteam.com/attorney-job-description

how artificial intelligence will affect law firms?

Documents